Hydrodynamic Flow in The Vicinity of A Nanopore Induced

NIH Public Access
Author Manuscript
Nanotechnology. Author manuscript; available in PMC 2014 June 21.
NIH-PA Author Manuscript
Published in final edited form as:

Nanotechnology. 2013 June 21; 24(24): 245202. doi:10.1088/0957-4484/24/24/245202.
Hydrodynamic flow in the vicinity of a nanopore induced by an

applied voltage
Mao Mao1, Sandip Ghosal2, and Guohui Hu3
Mao Mao: MaoMao2015@u.northwestern.edu
1Department
of Mechanical Engineering, Northwestern University, 2145 Sheridan Road,

Evanston, IL, 60208, USA
2Departments
of Mechanical Engineering and Engineering Sciences & Applied Mathematics,

Northwestern University, 2145 Sheridan Road, Evanston, IL, 60208, USA
3Shanghai
Institute of Applied Mathematics and Mechanics, Shanghai University, 149 Yanchang

Road, Shanghai 200072, P. R. China
Abstract
Continuum simulation is employed to study ion transport and fluid flow through a nanopore in a
solid-state membrane under an applied potential drop. Results show the existence of concentration
polarization layers on the surfaces of the membrane. The nonuniformity of the ionic distribution
gives rise to an electric pressure that drives vortical motion in the fluid. There is also a net
hydrodynamic flow through the nanopore due to an asymmetry induced by the membrane surface
charge. The qualitative behavior is similar to that observed in a previous study using molecular
dynamic simulations. The currentvoltage characteristics show some nonlinear features but are not
greatly affected by the hydrodynamic flow in the parameter regime studied. In the limit of thin
Debye layers, the electric resistance of the system can be characterized using an equivalent circuit
with lumped parameters. Generation of vorticity can be understood qualitatively from elementary
considerations of the Maxwell stresses. However, the flow strength is a strongly nonlinear
function of the applied field. Combination of electrophoretic and hydrodynamic effects can lead to
ion selectivity in terms of valences and this could have some practical applications in separations.
1. Introduction
Ionic conduction through nanometer-sized channels or pores is a common theme in

biological systems as well as in various manufactured materials such as membranes and
synthetic nanopores [1, 2, 3, 4, 5, 6, 7, 8, 9, 10]. Typical synthetic nanopore systems have a
length scale of a few nanometers to tens of nanometers. This is still sufficiently large
compared to molecular sizes that continuum simulations remain a powerful tool for studying
them [11, 12, 13, 14, 15, 16], even though, its accuracy decreases progressively as the pore
size approaches that of biological pores. This is because molecular dynamic (MD)
simulations at such mesoscales are computationally very expensive.
The objective of this paper is to study the hydrodynamic flow that arises in the vicinity of a
single cylindrical nanopore when a voltage is applied across the membrane. The role of
electrically induced hydrodynamic flows near pores, arrays of pores or semipermeable
membranes have been addressed by many authors [17]. Hydrodynamics is also implicated in
polymer translocation through nanopores where viscous resistance is shown to determine
translocation times [18, 19, 20] and stall forces [21] applied by optical traps to immobilize
DNA in nanopores against the electric driving force.
Mao et al.
Page 2
Under the appropriate set of circumstances, hydrodynamics at the entrance or the exit of a
nanopore has a strong influence on ion transport [17, 22, 23, 24]. Nanoslots and
nanochannels have been shown to exhibit ion selectivity similar to that of a permselective
membrane. Currentvoltage curves for such systems show a nonlinear saturation of the
current at high voltages followed by an overlimiting regime dominated by convective
transport by fluid vortices [24]. In the case of perm selective membranes, or equivalently for
a dense array of nanochannels in a membrane, the current saturation is caused by the loss of
carrier ions in the concentration polarization layer (CPL) adjacent to the membrane [25, 26,
27, 28]. The hydrodynamic flow has been attributed to a loss of stability in the CPL at
higher voltages [29, 30] that generates a vortex array and selects a diffusion length scale that
is different from the one in the quiescent Ohmic regime [31]. The overlimiting current is a
signature of this instability. These postulated micro-scale vortices have subsequently been
confirmed through direct observation [23, 32]. A related electrokinetic effect of relevance is
what has been termed electroosmosis of the second kind by Dukhin [33]. Here the flow is
driven by the field component tangential to a curved surface whereas the space charge layer
(SCL) itself is created by the normal component of the field. In the case of an isolated
nanopore with strongly overlapping Debye layers this results in corner vortices due to a
fieldfocusing effect and the limiting current plateau in the current voltage characteristic is
eliminated [34]. Such corner vortices are also expected at corners in micro-channels due to
leakage of the electric field into the dielectric [35].
In our previous paper [36] we utilized molecular dynamics (MD) simulation to study the
transport and flow field in a graphene sheet nanopore. We observed concentration
polarization, corner vortices and nonlinear currentvoltage relation for the graphene sheet
nanopore system. The results are in qualitative agreement with physical expectations and
theory. In this paper, we utilize continuum level simulation for transport and fluid flow in a
similar nanopore system. Such an approach has the advantage of computational speed but
certain features, for example steric effects, cannot be captured. We are interested in the
overlap between the continuum and MD approaches which we study by comparing our
simulations to the previous MD results. The rest of the paper is organized in this way: in
section 2, we present our modeling and describe our numerical approach. The results are
presented in section 3. Analysis and discussions are in section 4. In section 5 we provide
some concluding remarks.
2. Method
2.1. Problem formulation
Our nanopore system consists of a cylindrical pore of radius R drilled on a membrane of

thickness L. A cylindrical coordinate system is used. The pore connects two reservoirs that
have equal radius and depth, LR. A sketch of the nanopore system is shown in Figure 1. The
membrane surface is indicated by the lines: CD, DE and EF.
The bath is taken as a KCl solution where the salt is fully dissociated. The concentration of
either ionic species is taken as c0. We will assume, for simplicity, that the membrane has a
fixed surface charge of density and ignore any changes induced by shifts of the ionization
equilibrium (from various effects such as changes in salt concentration and pH). For the
purpose of estimation, a rule of thumb is that cannot exceed the value of one electron
charge per Bjerrum length [17]. Due to attraction of counterions, in this case K+, an electric
double layer (EDL) forms along the charged surfaces (CD, DE and EF). An external electric
field is used to generate current through the nanopore. The field will induce its own space
charge along surfaces due to concentration polarization (CP) thereby modifying the existing
EDL. The action of the tangential electric field on the space charges drags the fluid along the
membrane resulting in electroosmotic flow (or EOF). The Nernst-Planck-Poisson (NPP)
Mao et al.
Page 3
Stokes system of equations describes the coupling between ion transport, electric field and
hydrodynamics.
Assuming that a stationary state has been reached, the NPPStokes system of equations in
the solution domain consists of the continuity equation:
(1)
the Stokes equation:

(2)
the NernstPlanck equations for cations and anions:

(3)
(4)
and Poisson equation for the electric potential:
(5)
The fluid flow is governed by the Stokes equation with an additional electrostatic volume
force term, since the inertia term in the NavierStokes equation is inconsequential at the
nanoscale. Here u = ur + wz is the flow velocity, r and z being the unit vectors in the
radial (r) and axial (z) directions respectively, p is the fluid pressure and ci (i = 1, 2) is the
concentration of cation (K+) and anion (Cl). The electric potential is denoted by V and , ,
are the density, dynamic viscosity and permittivity of the solution respectively. The
diffusivity of the ith species is Di and zi is the valence (z1 = 1 and z2 = 1 for KCl). The flux
of species i is denoted by Ni and consists of contributions from diffusion, electrophoresis
and convection. The ion mobility (ui) in the NernstPlanck model obeys the Einstein
relation [37, 38] ui = Di/(kBT) where kB is the Boltzmann constant and T is the absolute
temperature, taken to be 300 K. In the membrane domain, we solve the Laplace equation,
(6)
where s is the permittivity of the membrane material.

The boundary conditions for the system are as follows: assuming the reservoirs have a much
larger volume than the pore, at BC and FG, the concentration gradient ci/r, the radial
velocity u and the radial electric field V/r are set to zero. At AB and GH, V = V/2
respectively and ci is set equal to the bulk concentration c0 while fluid pressure p is set to a
constant value (P0). At the membrane surfaces CD, DE and EF a nopenetration condition is
used for ions while the noslip condition applies for the flow. Applying the Gauss theorem
in an infinitely thin control volume across the surface, we get the jump condition for normal
component of the electric field (En) which is related to the surface charge density () as
[En] = . The tangential component of the field (Et) is continuous, [Et] = 0. Here the
brackets [ ] indicate the jump in the quantity enclosed across the membrane.
We use the finite element package COMSOL MULTIPHYSICS to solve the NPPStokes
system numerically. The triangular mesh is selected with refinement near the membrane
Mao et al.
Page 4
surface CD, DE and EF to ensure that the Debye layer is well resolved. The refinement
parameter is set such that the maximum mesh size in this region never exceeds one sixth the
Debye length. The results have been checked to ensure mesh independence. Small local
fillets are used at D and E to remove the sharp corners. In a related paper by Yossifon et al
[22] finite element simulations are used to study a set up very similar to ours. However, they
decouple the electrostatics and ion transport problems from the fluid flow by neglecting the
convective flux in their ion transport equations. We do not introduce such an approximation.
2.2. Dimensionless form
Results in this paper are reported in terms of dimensional quantities in order to facilitate
correspondence with the experimental literature. Nevertheless, we present the dimensionless
form of the equations here in order to identify the principal dimensionless parameters that
govern the physical behavior. We use the bulk concentration c0 as the scale for ci. Electric
potential is scaled with kBT/e. The reference length scale is chosen as the pore radius, R.
Normalizing the velocity, u, with U0 = D1/R and pressure, p, with U0/R, we get the
dimensionless form of the governing equations:
(7)
(8)
(9)
(10)
where = D/R and = (c0kBTR2)/(D1), where D, the Debye length, is defined as

. All the dimensionless variables and derivatives are superscripted with
*.
Two dimensionless parameters emerge. The first, = D/R is the ratio between Debye
length and the length scale R. It is an indication of how thin the Debye layer is in
comparison with the geometric scale. If is small, meaning that the Debye layer is thin
relative to the geometric length scale, the right hand side of the Poisson equation (10) will be
close to zero implying that the electrolyte is close to being electroneutral. Conversely, if ~
1, as is expected in most nanofluidic systems, Debye layers would overlap within the
nanopore. The second dimensionless parameter is
(11)
where um,1 is the mobility of cations. Multiplying by the factor eE0 in both the numerator
and denominator (E0 representing the strength of the applied electric field)
(12)
Mao et al.
Page 5
Here um,1eE0 is the velocity a K+ ion acquires under E0, or the electrophoretic velocity. The
quantity c0eR2E0/ also has a dimension of velocity and can be interpreted as the velocity
the fluid acquires under E0 because of the space charge density it carries, or an
electroosmotic velocity. Thus, is a ratio between the two velocities representing
electrophoretic and electroosmotic effects. It could also be related to the electric Reynolds
number RE = 1, defined as the ratio of the time scale of charge convection by flow to a
charge relaxation time set by Ohmic conduction [39]. For R = 2.5 nm and c0 = 0.1 M,
0.4 and 0.9, indicating that both finiteness of the Debye length and convective transport
needs to be taken into account.
3. Results
The results of the simulation depend on the geometry of the system (R, L and LR), the bulk
salt concentration c0, the membrane surface charge density and applied voltage bias V.
In this paper we fix the geometry of the nanopore system. We simulate a system
corresponding to R = 2.5 nm and L = 5 nm. LR is set to be 10L to make sure it is large
enough compared to the pore. The bulk concentration c0, surface charge density and
external voltage V are varied. Other physical properties are as follows. The relative
permittivity of fluid and membrane domains are respectively 80 (for water) and 3.9 (for
silica). Ion diffusivity D1 = 1.95 109 m2/s while D2 = 2.03 109 m2/s. Table 1 lists all
the parameters used in our simulations.
3.1. Concentration, electric field and flow field
The concentration distribution, electric field and flow field are shown in Figures 2,3 and 4
for some typical parameters. The figures correspond to a surface charge =-0.01C/m2, V
=0.4 V and c0 =0.1 M. The surface charge density of silica in electrolytes could vary from 0
to -0.1 C/m2 under different conditions [40].
The concentration distribution (Figure 2) for K+ and Cl are determined by both, the applied
field as well as the density of fixed charges on the membrane. If the membrane had no
surface charge, the K+ ions would show an enhanced density on the z > 0 side of the
membrane and a depletion on the z < 0 side. The negatively charged Cl ions would show
the opposite trend, so that the two Debye layers essentially form a parallel plate capacitor
with a small leakage of current through the pore. On the other hand, in the absence of an
applied external voltage, the membrane surface charge would form a Debye layer consisting
primarily of counterions. The polarity of the space charges along the membrane surfaces are
determined by these two competing effects. Due to the existence of a negative surface
charge on the membrane, we expect the concentration of K+ ions on the z > 0 side to be
enhanced relative to that of Cl ions. This is indeed what is observed in Figure 2(b), 2(c).
The relative importance of the membrane surface charge can be measured by the
dimensionless parameter * = L/(s V) which represents the ratio of the intrinsic charge
on the membrane to the surface charge induced by the applied field. For the situation shown
in Figure 2, * 0.1765.
The potential distribution in the vicinity of the pore is shown in Figure 3. Clearly most of the
potential drop occurs across the membrane/pore area, as the pore has a much larger
resistance than the electrolyte solution. Thus, the electric field is the strongest in the pore
region and there is a strong convergence of the electric field lines towards the pore.
However, there is also a leakage of field lines into the dielectric membrane due to the fact
that the conductivity of the solution is not infinite, though the field near the surface is
primarily radial. The local electric field is strongly affected by the surface charges, as can be
seen from the field lines near the membrane on the z < 0 side.
Mao et al.
Page 6
The radial component of the electric field acting on the thin layer of induced charges, cause
ions to move radially along the field creating electroosmotic flow. In the case depicted in
Figure 4, this flow is directed inwards (towards the pore) on the z > 0 side of the membrane
and outward (away from the pore) on the z < 0 side. Due to the large hydrodynamic
resistance the pore cannot accommodate this large radial influx and thus, mass conservation
drives an axial jet away from the pore as seen in Figure 4. The combination of the radial
flow along the surface with the axial jet produces a toroidal vortex near the periphery of the
pore. The existence of such a corner vortex in the reservoir near the entrance of a micro/
nano channel was discussed by Yossifon et al. [23, 22]. We also reported these vortices in
our study of the graphene nanopore using MD simulations [36]. The flow vorticity which is
entirely in the azimuthal direction is depicted in Figure 4(b). The vorticity in the Debye
layer adjacent to the membrane is many orders of magnitude stronger than the toroidal
vortex induced in the reservoir and is in the opposite direction. This Debye layer vorticity is
not shown in Figure 4(b) for clarity. Also note that the vorticity scale is logarithmic.
3.2. Electrical characteristics of the nanopore
The passage of ions through the nanopore, generates a detectable electric current. The
currentvoltage relation of the nanopore at different bulk concentrations c0 (and thus
different Debye lengths) is shown in Figure 5. The surface charge density is held fixed at
=-0.01 C/m2.
From Figure 5(d) it is seen that the nanopore has an Ohmic behavior when the bulk ion
concentration is high, c0 = 1 M. This is because at high bulk concentrations the Debye
length is thin and the system can be regarded essentially as an electroneutral uniform Ohmic
conductor. In this limit, the system may be regarded as a series connection of three separate
resistors representing (i) the bulk resistance of the bath (ii) the access resistance of the pore
and (iii) the pore resistance [41, 42]: Rtotal = Rbulk + Raccess + Rpore. The bulk resistance is
and the access resistance is [43] Raccess/2 = b/(4R). Here b
is the resistivity of the solution, which is related to the mobilities of ions, as
.
The pore resistance is the electrical resistance (Rc) of the electrolyte within the electroneutral bulk of the cylindrical pore in parallel with a second resistance (Rs) due to the
surface conductance of the polarized Debye layer. Thus, Rpore = Rc Rs/(Rc + Rs) where Rc =
(bL)/( R2). To calculate the surface resistance we assume that the polarized region
adjacent to the surface contains only counterions and the pore as a whole is electroneutral.
Then, if the electrophoretic mobility of K+ is K, the surface resistance can be calculated as:
Rs = L/(2 Rk||). Substituting the relevant numerical values, the resistance of the
nanopore system when c0 =1 M is found to be Rtotal = 3.09107 . The resistance calculated
directly from the slope of the I-V curve in Figure 5(d) is Rtotal =2.89 107 , in accord with
the simple lumped parameter model described above. Recently, Garaj et al [44] provided
conductance measurements on nanopores with diameters ranging from 5 to 23 nm in a
graphene sheet. They reported good agreement with calculations based on a model that
regarded the electrolyte as a homogeneous conductor. Our finding that the pore conductance
at 1 molar salt may be calculated from the Ohmic model is in accord with their observations.
At lower ionic concentrations the IV curves in Figure 5(a)-(c) show nonlinear behavior.
However, the nonlinearity is weak and does not show the classic behavior of a limiting
current leading to an over limiting current with increasing voltages as found in perm
selective membranes. This may seem surprising at first because a system of densely packed
nanopores is indeed expected to behave like a perm selective membrane. The difference can
Mao et al.
Page 7
be attributed to the radial inflow of ions that prevents the formation of a fully depleted CPL
next to the membrane [22].
3.3. Hydrodynamic characteristics of the nanopore
The hydrodynamic flow is driven by the electrical forces acting on the space charge adjacent
to the membrane. The space charge is induced by the fixed charge on the membrane as well
as by the normal component of the applied electric field. Electroosmotic flow due to the
former is expected to be a linear function of the applied voltage, however, the latter
generates an electroosmotic flow of the second kind in the terminology of Dukhin [33] and
has a nonlinear dependence on the voltage. As shown in Figure 4, the fluid is driven radially
along the membrane surface. Only a fraction of this radial flux crosses the membrane, the
remainder being deflected into an axial jet perpendicular to the membrane. In the situation
where the membrane has fixed charges, its Debye layer is enhanced on one side of the
membrane and depleted on the other side by the space charge layer created by the applied
field. Depending on the polarity and strength of the applied field, even the sign of the charge
in the membranes intrinsic Debye layer could reverse (refer to supplementary material).
Thus, the strengths of the axial jets on the two sides of the membrane would generally be
different for a charged membrane and there could be a net fluid flux through the pore. Such
a net flow however should not appear for a membrane with no surface charge. The fluid flux
through the pore was calculated by integrating the velocity over the pore area and is shown
as a function of the applied voltage in Figure 6. As expected, (a) the flux is always zero for
an uncharged membrane (b) the flux is a nonlinear function of the applied voltage but may
be regarded as approximately linear for weak applied voltages (c) the flux is an odd function
of the applied voltage. The last observation may be explained very simply by the fact that
the transformation V V is equivalent simply to a reversal of the z-axis and leaves
the physical system unaffected. Figure 6 also shows that the Q V characteristics are only
weakly affected by the electrolyte concentration in the bath (c0) which sets the Debye length
D. This is because the vortex in the reservoir is determined by the strength of the
Helmholtz-Smoluchowski slip velocity at the edge of the Debye layer and the slip velocity
itself is independent of the Debye length. In the context of Figure 4(b), changes in the Debye
layer thickness would change the width of the blue zone but leaves the toroidal vortex
largely unaffected.
3.4. Hydrodynamics induced ion selectivity
Many biological nanopores display ion selectivity beyond what might be expected solely
based on pore size [45]. The presence of a net hydrodynamic flow through a nanopore in a
charged membrane could result in an ion selectivity due to the competing effects of
electrophoresis and the convective flux due to the flow. This is because only the former is
sensitive to the ionic charge.
To demonstrate this effect we introduced a third charged species (concentration c3) of a
different valence z3 into the bottom reservoir. The diffusivity of the added species is taken
as D3 = 4.6 1011 m2/s, which is in the typical range of values seen in biological
macromolecules and polymers. The concentration of c3 is held fixed at c3 = 0.001 M at the
lower boundary and is set to zero at the upper boundary. Since this concentration is very low
compared to the bulk concentration of the carrier electrolytes (0.1 M), it can be considered a
trace species and its effect on the electric field and hydrodynamic flow neglected. Thus, we
evolved c3 as a passive scalar using the previously computed electric field and fluid velocity.
The remaining parameters in the simulation are V = 0.5 V and = 0.01 C/m2. The flux
of the added species through the pore is shown in Figure 7 as a function of the ion valence
z3. The flux is of course zero if z3 > 0 as cations are driven away from the pore and diffusive
flux is negligible. However, for anions, due to the competing effects of electrophoresis and
Mao et al.
Page 8
convection, ions with valence in the range 0 > z3 > 1 are also excluded. If an array of
nanopores is fabricated to bridge a pair of parallel microfluidic channels, the pores will
behave as a permselective membrane blocking all ions for which z3 > 1. Such a system
could therefore be used to separate biomolecules, for example proteins, using essentially the
same principle as is used in the desalination of water [46]. Furthermore, the permselective
properties can be modulated by changing the voltage applied between the microchannels. A
similar effect was demonstrated by Karnik et al [47] to make a gate controlled nanofluidic
transistor though they attributed the selective properties to the Donnan potential in the
channel.
4. Discussion
In this paper, we present a continuum level description of a problem that we had studied
previously using MD simulation [36], namely, hydrodynamics and ion transport through a
nanopore in a membrane. We find qualitative agreement of the two simulations: similar
concentration polarization, corner vortices, nonlinear currentvoltage characteristics as well
as net flow through the nanopore. However, in the present paper, it is the membrane surface
charge that creates the asymmetry responsible for the net flow. In our MD simulation, the
membrane had no charge, however, the asymmetric space charge was due to a difference in
mobilities between sodium and chlorine ions. Here we use potassium and chlorine ions, that
have almost identical mobilities. Thus, a net flow can be generated by any condition that
induces an asymmetry between the two sides of the membrane.
There are however some effects that are only captured in the MD simulation but not in our
continuum simulation. One such effect is the clearly observable increase in density of water
due to preferential alignment of the water molecules in the large electric field close to the
pore. Since the continuum simulations are based on the incompressible fluid equations, this
effect is a priori excluded. The ionic distributions in the MD simulations show a peak a few
ionic radii from the surface. This is due to a steric effect: ions cannot get closer than an ionic
radius of the surface. Clearly this effect is not observable in the continuum version.
The transport of ionic species is governed by the NernstPlanck equation. The total flux
consists of the diffusive flux, electrophoretic flux and electroosmotic flux. The relative
importance of electrophoresis and electroosmosis can be determined by comparing the
electrophoretic velocity and electroosmotic velocity. Both velocities reach their peak value
within the pore region where the electric field is largest. In the present study the peak
electrophoretic velocity is ep 7.7 m/s. From Figure 4, the maximum electroosmotic
velocity is only about 0.2 m/s. Thus, transport of ions in the pore is dominated by
electrophoresis not electroosmosis as argued by Chang et al. [17]. The vortical flow could
however affect the current-voltage characteristics indirectly [22, 17, 30] by modifying the
ionic distributions at the pore entrance. In the current study, this effect is not of significant
magnitude. The vortical flow induced in the bath could nevertheless have other significant
effects. It could compete with electrophoretic transport of low mobility species such as
macromolecules, as seen in Figure 7. Another possibly significant effect that we did not
study in this paper is the translocation of polyelectrolytes across pores, a very important
biological process [48] that has also received much attention as a biotechnology tool [49].
Even though the translocation time of the polymer is set by the electric field within the pore
and the pore geometry [18, 19, 20] the rate of capture of polymers by the pore can be greatly
enhanced by the toroidal vortex in the bath. This is because the polymer moves towards the
pore by diffusion and one end of the polymer must insert into the pore by the process of
reptation before the electric field can take hold and start the translocation. Both of these
processes are very slow. The vortex can convect the polymer towards the pore as well as
Mao et al.
Page 9
provide the shear needed for it to undergo a coil-stretch transition thus greatly helping the
process of insertion in the pore.
In conclusion, we can say that except for certain very specific features, most properties of
the nanopore system are adequately described by the continuum NPP-Stokes model for fluid
and ion transport. This is encouraging, because such continuum simulations are many orders
of magnitude cheaper than the corresponding MD simulations and therefore represent a
good compromise between accuracy and computational expenses at the mesoscale.
Supplementary Material
Refer to Web version on PubMed Central for supplementary material.
Acknowledgments
This work was supported by grant number R01HG004842 from the National Human Genome Research Institute,
National Institutes of Health. One of us (SG) acknowledges support from the Leverhulme Trust (UK).
References
1. Doyle, Declan A.; Cabral, Joo Morais; Pfuetzner, Richard A.; Kuo, Anling; Gulbis, Jacqueline M.;
Cohen, Steven L.; Chait, Brian T.; MacKinnon, Roderick. The structure of the potassium channel:
Molecular basis of K+ conduction and selectivity. Science. 1998; 280(5360):6977. [PubMed:
9525859]
2. Rhee, Minsoung; Burns, Mark A. Nanopore sequencing technology: research trends and
applications. Trends in Biotechnology. Dec; 2006 24(12):580586. [PubMed: 17055093]
3. Venkatesan, Bala Murali; Bashir, Rashid. Nanopore sensors for nucleic acid analysis. Nature Nano.
Oct; 2011 6(10):615624.
4. Branton, Daniel; Deamer, David W.; Marziali, Andre; Bayley, Hagan; Benner, Steven A.; Butler,
Thomas; Di Ventra, Massimiliano; Garaj, Slaven; Hibbs, Andrew; Huang, Xiaohua; Jovanovich,
Stevan B.; Krstic, Predrag S.; Lindsay, Stuart; Ling, Xinsheng Sean; Mastrangelo, Carlos H.;
Meller, Amit; Oliver, John S.; Pershin, Yuriy V.; Ramsey, J Michael; Riehn, Robert; Soni, Gautam
V.; Tabard-Cossa, Vincent; Wanunu, Meni; Wiggin, Matthew; Schloss, Jeffery A. The potential and
challenges of nanopore sequencing. Nature Biotech. 2008; 26(10):11461153.
5. Roux, Benot; Allen, Toby; Bernche, Simon; Im, Wonpil. Theoretical and computational models of
biological ion channels. Quarterly Reviews of Biophysics. 2004; 37(01):15103. [PubMed:
17390604]
6. Demming, Anna. Nanoporesthe holy grail in nanotechnology research. Nanotechnology. Jun;
2012 23(25):250201250201. [PubMed: 22653269]
7. Liu, Ling; Chen, Xi; Lu, Weiyi; Han, Aijie; Qiao, Yu. Infiltration of Electrolytes in Molecular-Sized
Nanopores. Physical Review Letters. May.2009 102(18):184501. [PubMed: 19518875]
8. Qiao, Yu; Liu, Ling; Chen, Xi. Pressurized liquid in nanopores: a modified Laplace-Young
equation. Nano letters. Mar; 2009 9(3):9848. [PubMed: 19182906]
9. Lu, Weiyi; Kim, Taewan; Han, Aijie; Qiao, Yu. Electrically controlling infiltration pressure in
nanopores of a hexagonal mesoporous silica. Materials Chemistry and Physics. Mar; 2012 133(1):
259262.
10. Kong, Xinguo; Qiao, Yu. An electrically controllable nanoporous smart system. Journal of Applied
Physics. 2006; 99(6):064313.
11. Ai, Ye; Qian, Shizhi. Direct numerical simulation of electrokinetic translocation of a cylindrical
particle through a nanopore using a Poisson-Boltzmann approach. Electrophoresis. Apr; 2011
32(9):9961005. [PubMed: 21455912]
12. Ai, Ye; Qian, Shizhi. Electrokinetic particle translocation through a nanopore. Physical chemistry
chemical physics : PCCP. Mar; 2011 13(9):406071. [PubMed: 21229154]
Mao et al.
Page 10

13. Ai, Ye; Zhang, Mingkan; Joo, Sang W.; Cheney, Marcos A.; Qian, Shizhi. Effects of
Electroosmotic Flow on Ionic Current Rectification in Conical Nanopores. Journal of Physical
Chemistry C. 2010:38833890.
14. Stein, Derek; Deurvorst, Zeno; van der Heyden, Frank HJ.; Koopmans, Wiepke JA.; Gabel, Alan;
Dekker, Cees. Electrokinetic Concentration of DNA Polymers in Nanofluidic Channels. Nano
Letters. Feb; 2010 10(3):765772. [PubMed: 20151696]
15. Lee, Choongyeop; Joly, Laurent; Siria, Alessandro; Biance, Anne-Laure; Fulcrand, Rmy;
Bocquet, Lydric. Large apparent electric size of solid-state nanopores due to spatially extended
surface conduction. Nano letters. Aug; 2012 12(8):403744. [PubMed: 22746297]
16. Liu, Hui; Qian, Shizhi; Bau, Haim H. The effect of translocating cylindrical particles on the ionic
current through a nanopore. Biophysical journal. Feb; 2007 92(4):116477. [PubMed: 17142291]
17. Chang, Hsueh-Chia; Yossifon, Gilad; Demekhin, Evgeny A. Nanoscale Electrokinetics and
Microvortices: How Microhydrodynamics Affects Nanofluidic Ion Flux. Annual Review of Fluid
Mechanics. 2011
18. Ghosal, Sandip. Electrophoresis of a polyelectrolyte through a nanopore. Physical Review E. 2006;
74(4):41901.
19. Ghosal, Sandip. Effect of salt concentration on the electrophoretic speed of a polyelectrolyte
through a nanopore. Physical Review Letters. 2007; 98(23):238104. [PubMed: 17677940]
20. Ghosal, Sandip. Electrokinetic-flow-induced viscous drag on a tethered DNA inside a nanopore.
Physical Review E. 2007; 76(6)
21. van Dorp, Stijn; Keyser, Ulrich F.; Dekker, Nynke H.; Dekker, Cees; Lemay, Serge G. Origin of
the electrophoretic force on DNA in solid-state nanopores. Nature Physics. Mar; 2009 5(5):347
351.
22. Yossifon, Gilad; Mushenheim, Peter; Chang, Hsueh-Chia. Controlling nanoslot overlimiting
current with the depth of a connecting microchamber. Europhysics Letters. 2010; 90(6):64004.
23. Yossifon, Gilad; Chang, Hsueh-Chia. Selection of Nonequilibrium Overlimiting Currents:
Universal Depletion Layer Formation Dynamics and Vortex Instability. Physical Review Letters.
2008; 101(25):254501. [PubMed: 19113713]
24. Chang, Hsueh-Chia; Yossifon, Gilad. Understanding electrokinetics at the nanoscale: A
perspective. Biomicrofluidics. 2009; 3(1)
25. Levich, Veniamin G. Physicochemical hydrodynamics. New York: Prentice Hall; 1962.
26. Rubinstein, Isaak; Shtilman, Leonid. Voltage against current curves of cation exchange
membranes. Journal of the Chemical Society, Faraday Transactions 2: Molecular and Chemical
Physics. 1979; 75:231246.
27. Ben Y, Chang Hsueh-Chia. Nonlinear Smoluchowski slip velocity and micro-vortex generation.
Journal of Fluid Mechanics. 2002; 461:229238.
28. Yariv, Ehud. Asymptotic current-voltage relations for currents exceeding the diffusion limit.
Physical Review E. Nov.2009 80(5)
29. Rubinstein, Isaak; Zaltzman, Boris. Electro-osmotically induced convection at a permselective
membrane. Physical Review E. 2000; 62(2):22382251.
30. Zaltzman, Boris; Rubinstein, Isaak. Electro-osmotic slip and electroconvective instability. Journal
of Fluid Mechanics. 2007; 579:173226.
31. Yossifon, Gilad; Mushenheim, Peter; Chang, Yu-Chen; Chang, Hsueh-Chia. Nonlinear currentvoltage characteristics of nanochannels. Physical Review E. 2009; 79(4):46305.
32. Rubinstein, Shmuel M.; Manukyan, Gor; Staicu, Adrian; Rubinstein, Issac; Zaltzman, Boris;
Lammertink, Rob; Mugele, Frieder; Wessling, Matthias. Direct Observation of a Nonequilibrium
Electro-Osmotic Instability. Physical Review Letters. 2008; 101(23):236101. [PubMed:
19113567]
33. Dukhin, Stanislav S. Electrokinetic phenomena of the second kind and their applications.
Advances in Colloid and Interface Science. Mar; 1991 35(0):173196.
34. Yossifon, Gilad; Mushenheim, Peter; Chang, Yu-Chen; Chang, Hsueh-Chia. Eliminating the
limiting-current phenomenon by geometric field focusing into nanopores and nanoslots. Physical
Review E. 2010; 81(4):46301.
Mao et al.
Page 11

35. Yossifon, Gilad; Frankel, Itzchak; Miloh, Touvia. On electro-osmotic flows through microchannel
junctions. Physics of Fluids. 2006; 18(11):117108.
36. Hu, Guohui; Mao, Mao; Ghosal, Sandip. Ion transport through a graphene nanopore.
Nanotechnology. 2012
37. Einstein, Albert. ber die von der molekularkinetischen theorie der wrme geforderte bewegung von
in ruhenden flssigkeiten suspendierten teilchen. Annalen der Physik. 1905; 322(8):549560.
38. Ashcroft, Neil W.; Mermin, N David. Solid state physics. New York: Holt, Holt, Rineheart and
Winston; 1988. p. 826
39. Feng, James Q. Electrohydrodynamic behaviour of a drop subjected to a steady uniform electric
field at finite electric reynolds number. Proceedings of the Royal Society of London Series A:
Mathematical, Physical and Engineering Sciences. 1999; 455(1986):22452269.
40. Rodrigues, Flvio A.; Monteiro, Paulo JM.; Sposito, Garrison. Surface charge density of silica
suspended in wateracetone mixtures. Journal of Colloid and Interface Science. 1999; 211(2):408
409. [PubMed: 10049558]
41. Kowalczyk, Stefan W.; Tuijtel, Maarten W.; Donkers, Serge P.; Dekker, Cees. Unraveling SingleStranded DNA in a Solid-State Nanopore. Nano Letters. Mar; 2010 10(4):14141420. [PubMed:
20235508]
42. Smeets, Ralph MM.; Keyser, Ulrich F.; Krapf, Diego; Wu, Meng-Yue; Dekker, Nynke H.; Dekker,
Cees. Dependence of Ion Transport and DNA Translocation through Solid-State Nanopores. Nano
Letters. Dec; 2005 6(1):8995. [PubMed: 16402793]
43. Hall, James E. Access resistance of a small circular pore. The Journal of General Physiology.
1975; 66
44. Garaj S, Hubbard W, Reina A, Kong J, Branton D, Golovchenko JA. Graphene as a subnanometre
trans-electrode membrane. Nature. 2010; 467(7312):190193. [PubMed: 20720538]
45. Hille, Bertil. Ion channels of excitable membranes. Sunderland, Mass: Sinauer Associates, Inc;
1984.
46. Oren, Yoram; Rubinstein, Isaac; Linder, Charles; Saveliev, Galina; Zaltzman, B.; Mirsky, Elena;
Kedem, Ora. Modified heterogeneous anion-exchange membranes for desalination of brackish and
recycled water. Environmental Engineering Science. Nov-Dec;2002 19(6):513529.
47. Karnik, Rohit; Fan, Rong; Yue, Min; Li, Deyu; Yang, Peidong; Majumdar, Arun. Electrostatic
control of ions and molecules in nanofluidic transistors. Nano Letters. 2005; 5(5):943948.
[PubMed: 15884899]
48. Shari, Karim; Ghosal, Sandip; Matouschek, Andreas. The force exerted by the membrane potential
during protein import into the mitochondrial matrix. Biophysical Journal. 2004; 86(6):36473652.
[PubMed: 15189861]
49. Bell, Nicholas AW.; Engst, Christian R.; Ablay, Marc; Divitini, Giorgio; Ducati, Caterina; Liedl,
Tim; Keyser, Ulrich F. Dna origami nanopores. Nano Letters. 2012; 12(1):512517. [PubMed:
22196850]

Mao et al.
Page 12

Figure 1.
A sketch of the nanopore system showing cylindrical coordinates. The system is

axisymmetric.

Mao et al.
Page 13

Figure 2.
Equilibrium ionic distributions in the nanopore system. The parameters are: =-0.01 C/m2
V =0.4 V and c0 =0.1 M.
(a) The steady state concentration distribution (normalized by the bath concentration, c0) in
the nanopore system. The system is axisymmetric. For convenience, the left panel shows the
concentration distribution of potassium ions, K+ and the right panel is used for chloride ions,
for Cl.
(b) Normalized ion concentrations along line r = 0 (pore center line)
(c) Normalized ion concentrations along line r = 30 nm (faraway from pore)
Mao et al.
Page 14

Figure 3.
The electric potential distribution in the nanopore system. Parameters are, =-0.01 C/m2
V =0.4 V and c0 =0.1 M.
(a) The electric potential distribution. The numerical values in the color map are in Volt and
field lines are shown in white.
(b) Distribution of the electric potential along the zaxis.
(c) The radial distribution of the electric potential along the upper and lower membrane
surface (z = L/2).
Mao et al.
Page 15

Figure 4.
Hydrodynamic flow in the nanopore system. Parameters are =-0.01 C/m2 V =0.4 V and
c0 =0.1 M.
(a) Flow field in the nanopore system. The color scale indicates magnitude of velocity in m/
s. White lines are streamlines and arrows indicate the flow direction.
(b) Vorticity in the azimuthal direction in units of 1/s. Note that the scale is logarithmic. The
vorticity due to the shear in the Debye layer is indicated as a flat blue color and is not
represented by the color scale. The vorticity in this zone is in the opposite direction and is
much larger than the vorticity in the bath represented by the color scale.
Mao et al.
Page 16

Figure 5.
Currentvoltage characteristics of the nanopore system for different bulk concentrations: (a)
c0 =0.001 M (b) c0 =0.01 M (c) c0 =0.1 M (d) c0 = 1.0 M. The membrane charge is fixed at
=-0.01 C/m2.

Mao et al.
Page 17

Figure 6.
Flow rate through the nanopore at fixed surface charge (left) and fixed salt concentration
(right).
(a) Total flow rate through nanopore when =-0.01 C/m2
(b) Total flow rate through nanopore when c0 =0.1 M

Mao et al.
Page 18

Figure 7.
The dependence of the various flux contributions to the total flux for a charged trace species
as a function of its valence (z3). When z3 < 0, the convective and electrophoretic
components of the flux are comparable but in opposite directions. Here c0 =0.1M, =-0.01
C/m2 and V =0.5 V.

L(nm)
R(nm)
2.5
Geometry
1.0
0.1
0.01
0
-2.0 to 2.0
0.01
(C/m2)
-0.01
V(V)
0.001
c0(M)
Physical
3.9(water)
80(silica)
Relative permittivity
Parameters used in the simulations
2.03(Cl)
1.95(K+)
Diffusivity (109 m2/s)
Table 1
Mao et al.
Page 19

Hydrodynamic Flow in The Vicinity of A Nanopore Induced

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Hydrodynamic Flow in The Vicinity of A Nanopore Induced

Uploaded by

Copyright:

Available Formats

NIH Public Access

NIH-PA Author Manuscript

Published in final edited form as: